We have been collaborating with Drs. David Roberts and Henry C. Krutzsch (Laboratory of Pathology, NCI) studying inhibition of endothelial cell proliferation by synthetic peptide analogs of the matrix protein, thrombospondin-1 (TSP-1). The major goal of this collaboration was to design and test therapeutically useful analogs of thrombospondin sequences that could block new blood vessel development (angiogenesis) necessary for the growth of primary and metastatic malignant tumors or causing other neovascular pathologies, such as angiogenic retinopathies. The role of this Section in these studies has been in assisting with the design of peptide analogs, such as retro-inverso forms, wherein natural L-amino acids are inverted to the D-isomers and the sequence direction is reversed. Also, we have been concerned with constructing polymer conjugates of these peptides as drug delivery vehicles and with the characterization of aggregated forms of the peptides. During this past year, animal model studies of the effect of TSP- 1 peptide analogs on growth of human breast cancer cells were completed. Immunodeficient, "nude", mice had been infected with human breast carcinoma cells (MDA-MB-435S). After an initial dormant period before tumor growth normally would begin, the animals were dosed with peptide intravenously each day for number of weeks and observed for a few weeks longer. Their tumor mass was then measured and compared with saline-injected controls. Results with a 17-residue retro-inverso sequence from the second type-1 repeat of TSP-1 (Ac-sswpswhswGGdqkfrk-NH2, where lower case letters stand for D-amino acids) showed only 12% of the tumor growth of controls. Another peptide analog in which two of the residues were altered gave 16% of control growth. In our section, we separated aggregated from nonaggregated peptide of a truncated, 13-residue analog (see last year?s annual report). Results from these trials showed 34% (aggregated) and 51% (nonaggregated) of the control growth. No toxic effects were reported and optimal dosing may not have been achieved. Yet, the results, especially with the retro-inverso natural sequence, are encouraging. The retro-inverso analogs represent a favorable alteration for potential therapeutic use, since these forms are not readily degraded by proteolytic enzymes in vivo. At the start of this reporting period, I prepared a conjugate of an 11-residue C-terminal peptide fromTSP-1 (synthesized by Dr. Krutzsch) by covalently linking it to a high molecular weight polysucrose molecule (Ficoll). This peptide contains a 9-residue sequence which, when linked to a solid surface serves as a ligand for CD47, reportedly a costimulatory receptor for the T-cell antigen receptor complex (CD3). Dr. Roberts tested the polyvalent soluble Ficoll conjugate for this activity and compared it with a control peptide, similarly conjugated. Unfortunately, no significant difference was observed between the two preparations. It was concluded that, possibly, the earlier observations on solid carriers resulted from the action of nonspecific hydrophobic forces.